Exploratory Analysis

The State of Our Climate

Authors

• Drew King kingdrew@uw.edu
• Derek J Liu deliu@uw.edu
• Sierra Briscoe sierrak9@uw.edu
• Alicia Lee leeee@uw.edu

Date

Summer 2023

Abstract

Providing a data visualization that presents a comprehensive investigation into the impact of climate change between the years 1750 and 2015, employing temporal and trend-line mapping techniques to explore the evolution of climate-related parameters, and provide discernment for the individual factors influencing climate change over time and their respective influence. The objective of this data visualization is to enhance understanding of the drivers of climate change, enabling more accurate projections of future climate scenarios. Considering the prevailing climate emergency, a motivation for this data visualization project is to provide valuable insights towards cultivating a more environmentally-conscious world.

Keywords

(Climate Change, Geographic Data, Atmospheric Science, Time series data)

Introduction

The question that this project will answer is how climate temperature has changed over time. Climate change is an important issue globally, and being able to understand the progression of climate change is important data for creating action. To address this question, a series of visualizations derived from a dataset will be created; a temporal map, showing a heatmap laid over a world map, a bubble chart that represents the difference in heat by year, an area graph representing temperature over time with a slider to alter the time interval, and a line chart that depicts the overall velocity of temperature change over time from 1750 to now, knowing that earth temperatures have been relatively stable before 1750 throughout recent human history. The velocity at which the temperature is changing overall is also of concern, because it can be used to set the trendline for making future predictions. As a result, there are many factors the data provides that are worth identifying such as:

• How has global temperature changed over time?
• What is the velocity of temperature change throughout time?
• Which continent/country/city experienced the greatest changes in temperature over time?
• Other questions worth considering while exploring this dataset include:
• What major historical events have taken place in this timeline that may have influenced the climate?
• What does the trend in temperature suggest about the future?

Finding answers to these questions is important since they will ensure scientific accountability, raise awareness in large audiences, maintain documentation of historical data, and hopefully advocate for international collaboration. It is important to consider that this data set has some limitations; the data dates back to 1750, all the way up to 2017. While this is a large dataset, it isn’t necessarily the most updated representation of the climate today since it excludes the previous six years of data. Therefore a trendline for future predictions may be helpful to use and compare with current temperatures for accuracy. This dataset lacks granular data for every place on the globe, a limitation of this dataset. The missing data is notable because it inhibits the accuracy of the data representation and it will be countered by supplying more refined data on the location and date of significant temperature changes over time by cities.

Related Work

The topic under investigation pertains to the temporal rate of climate temperature change. Given the contentious nature of climate change, a plethora of ‘information’ on its effects and causes exists. Accessing reliable sources such as academic journals or the platforms of esteemed scientific organizations like the UN, NOAA, or NASA is recommended.

• The UN provides a ‘Climate Action Note’, which provides data maps on the increase of emissions since 1990. This data set provides interactive data maps on climate change per total emissions and capita emissions, by country. What is most interesting about this particular data set is that they provide economic data on the countries contributing to climate change as well. This is beneficial in understanding the very real economic disparities between countries, and how that affects carbon emissions. For example, by analyzing the economic welfare of each individual country in the data set, the UN found that the poorest countries only make up half of carbon emissions. The top 1% of the globe represents more than twice these emissions. This information is useful in understanding how to disseminate resources and support to countries in need (Benjamin 2023).

• The NOAA provides a ‘Global Climate Dashboard’, which provides 18 different line graphs ranging from Greenhouse Gases, Sea Level, and Surface Temperature to Ocean Heat, Spring Snow, and sunlight by metric. There are two indicator options, filtered between ‘climate change’ and ‘natural variability’ which increases the confidence in correlation between the state of the climate crisis which accelerated at a rate that went beyond what was natural for that particular time period or environment.

• NASA provides data on the ‘Vital Signs of our Planet’, the data provides information on Carbon Dioxide (parts per million), Global Temperature (degrees in Celsius since the pre industrial period), Methane (parts per billion), Arctic Sea Ice Extent (min percent per decade since ’79), Ice sheets(by billion metric ton per year), Sea Level (inches since 1993), and Ocean Warming (zetta-joules since 1955). By extracting these indicators, NASA provides a closer look at the different natural elements that are affected, and how drastically they have changed over time. When opened, each data-set provides either a trend-line graph, a regular line graph, or a temporal world map that is interactive.

All of these articles provide additional supporting data and impact to the questions we are trying to answer, but more than that, they are all coming from major scientific organizations that are credible, reliable, and actively involved in providing systemic solutions to the climate crisis. The goal of this project is to create data visualizations which contribute to the accessibility of climate data information, in a way that is readable and understandable to an audience.

The Dataset

Dataset biography:

• This data was found on Kaggle. The data set can be found here: Climate Change Data Set

• The data was collected by Berkeley Earth and Kristen Sissener

Data collection methodology:

• There are several different ways this data was collected because the data dates all the way back to 1750, it is difficult to have consistent measurement. Early periods of temperature data were gathered using mercury thermometers, post 1940’s construction moved weather stations around, and the 1980’s introduced a device known to have a cooling bias; digital thermometers. There were 16 different data archives with a whopping 1.6 billion individual temperature records reported. Berkeley Earth divided the data into categories based on min, max, confidence intervals, global scale, by specific location, and by date.

• This data was collected to identify location indicators affecting climate temperature change over time. Authors Berkeley Earth and Kristen Sissener acknowledge the ongoing debate on whether climate change is a real issue, and collected this wide range of temporal data by location in order to represent a seemingly unbiased data set for analysis. By providing data based on location rather than population per capita, the data eradicates any room for socioeconomic debate and only provides space to analyze the correlation between date and location in temporal history.

Dimensions of dataset

Dimensions	global_temp	country_temp	city_temp
Rows	3192	577462	8599212
Columns	9	4	7

Some ethical questions considered when working with this data include…

• Is the data representative of the global population?

  • Considering that there is no information on each individual place across the world, the data will not provide very high confidence of causation or correlation between environmental placement and its impact. In order to do so, there would need to be granular data provided for each and every individual city, state, and country worldwide. Knowing that this data set spans across such a large amount of time, it is fair to assume this would not be a possibility to provide, however the ethical dilemma of representation within the data-set still very much stands.

• Does the data have a motive or an agenda behind it?

  • Knowing that climate change is not just a scientific problem, but also a highly politicized issue forces the consideration of integrity behind the data collection. What is being left out? How do the indicators selected create gaps in our data set? Are these gaps intentional or causal? It is important to consider these questions when approaching these data sets in order to consider the marginalized groups. Because we know that the poorest countries in the world lack resources that would provide them with environmental preservation efforts, it would be unethical to assume the baseline for each country remains the same.

It is important to consider that this data set has some limitations; the data dates back to 1750, all the way up to 2015. While this is a large data-set, it isn’t necessarily the most updated representation of the climate today since it excludes the previous six years of data. Therefore a trend-line for future predictions may be helpful to use and compare with current temperatures for accuracy. This data-set also lacks granular data for all surface locations on the globe, another limitation of this data-set. The missing data is notable because it inhibits the accuracy of the data representation.

Implications

When addressing the challenge of climate change, technologists, designers, and Policymakers are the result of addressing tangible solutions for the future of climate change and how we understand its data. The fusion of innovative technologies with sustainable design that is supported and enforced by effective policies ensures the acceleration of the ways sustainable practices across the globe will ensure a more environmentally-conscious future.

• Technologists: Knowing the climate is in a state of emergency, technologists have the resources, and understanding, and will play a crucial role in developing sustainable solutions and green technologies. This includes renewable energy sources, energy-efficient infrastructure, climate monitoring systems, and advanced data analytics to inform decision-making.
• Designers: Sustainable design practices will be essential to mitigate the impact of climate change. Designers should prioritize eco-friendly materials, energy-efficient buildings, and resilient infrastructure to adapt to changing environmental conditions.
• Policymakers: Policymakers have the responsibility to implement effective climate policies, promote international cooperation, and enforce environmental regulations. This may include carbon pricing, emission reduction targets, and incentives for adopting clean technologies.

Limitations & Challenges

Possible limitations or problems when working with this particular dataset are concerns of the locations being represented. Knowing that this data set does not have granular time series data for each individual city, state, or country as it is aggregated monthly. It is important to consider how this presents difficulties when deciphering events with the highest or lowest impact on the climate. Alternatively, the information provided comes from a multitude of different measurement sources dating all the way back to the pre-industrial revolution. The ways in which temperature measurements are gathered today are different than they were in 1750, and the results of these measurements reflect that. The authors of this data set mention that they collected data from measurements based off of very dated instruments, such as mercury thermometers of the pre-industrial age, and the digital thermometers used in the late 1900’s, which are known to have a cooling bias.

These concerns question not only the integrity of the data’s accuracy and precision, but also the invisible power structures that it represents. It is clear not every location was capable of being held accountable for proper measuring techniques, therefore the data collected must consider those disparities. There is missing data for many countries and cities before 1850, there is even more missing data for Antarctica. Additionally, considering the main question to answer is how temperature of the climate has changed over time, the data limits or completely dismisses the ability to extract individual variables that also represent environmental indicators of significant climate change such as ocean heat levels or snow melt in the spring. This is important because these factors are crucial in determining temperature change over time, arguably more so than analyzing indicators based on location.

Summary Information

Between the years 1750 and 2015, the average global temperature has experienced a significant change, with an average increase of approximately 1.11°C. In 1752, the minimum average temperature was recorded at 5.78°C, while in 2015, the maximum average temperature reached 9.83°C. A notable temperature of 8.38°C was observed in 1906 as the median average temperature during this period. Finally, we can calculate the mean of these temperatures: Mean temperature = (1752 * 5.779833 + 2015 * 9.831 + 1906 * 8.379083) / (1752 + 2015 + 1906). This equation gives us a mean temperature finding of approximately 8.04 degrees Celsius. Overall, this means that given the values found in our data set, the overall mean temperature for the years ranging from 1750-2015 is roughly 8.04 degrees Celsius.

We found:

• Between 1750 to 2015 the average change in land temperature globally is: 1.111636^^0 C
• The lowest global land average temperature since 1750 was in 1752 at 5.779833^^0 C
• The highest global land average temperature since 1750 was in 2015 at 9.831000^^0 C
• The median temperature globally by year occurred in 1906 and is 8.379083^^0 C
• The average temperature between 1750 and 2015 was 8.369337^^0 C

Table

This table is included to give a more detailed understanding of our global data summary statistics, this table includes average land and sea temperatures as well as the confidence interval for each calculation.

Date	Event	LandTemp	TempConfidence	MaxTemp	MaxTempConfidence	MinTemperature	MinTempConfidence	LandAndSeaTemp	LandAndSeaTempConfidence
1750 - 2015	chg_avg_temp	1.111636	-2.5456515	20.90400	0.1065000	-1.518000	0.1417500	16.05858	0.0608333
1750 - 2015	avg_temp	8.369337	0.9457185	20.08746	0.4797816	-3.070632	0.4318489	15.21257	0.1285321
1752	min_avg_temp	5.779833	2.9770000	NA	NA	NA	NA	NA	NA
1906	med_avg_temp	8.379083	0.2749167	20.05800	0.3840000	-3.637000	0.3421667	15.03192	0.1285833
2015	max_avg_temp	9.831000	0.0921667	20.90400	0.1065000	-1.518000	0.1417500	16.05858	0.0608333

Chart 1

Include a chart

Describe why you chose this chart and what information it reveals

Here’s an example of how to run an R script inside an RMarkdown file:

Chart 2

Include a chart

Describe why you chose this chart and what information it reveals

Chart 3

Include a chart

Describe why you chose this chart and what information it reveals

Citations

Benjamin, A. (2023, January 16). Climate Action - UN. Climate Action: What We Do. Retrieved July 26, 2023, from https://www.unep.org/explore-topics/climate-action/what-we-do/climate-action-note/state-of-climate.html?gclid=Cj0KCQjwiIOmBhDjARIsAP6YhSW-htOedd08vlz5LUT105nJ-_fRdkCJfswUJCBBTe8c4fdD_amOjzkaAkGWEALw_wcB

(NASA): Climate Change: Vital Signs of the Planet. (n.d.). Home. Retrieved July 26, 2023, from https://climate.nasa.gov

(NOAA): Global Climate Dashboard | NOAA Climate.gov. (n.d.). Climate.gov. Retrieved July 26, 2023, from https://www.climate.gov/climatedashboard